the materiality of human capital to corporate financial ......human capital research has been...

by Aaron Bernstein and Larry Beeferman

April 2015

The Materiality of Human Capital to Corporate

Financial Performance

This research was funded by the Investor Responsibility Research Center Institute (IRRCi) and co-authored by Larry Beeferman and Aaron Bernstein with the Labor and Worklife Program at Harvard Law School. The analysis, opinions and perspectives herein are the sole responsibility of the authors.

The copyright for this report is held by the President and Fellow of Harvard College. The material in this report may be reproduced and distributed without advance permission, but only if attributed. If reproduced substantially or entirely, it should include all copyright and trademark notices.

© Copyright 2015, President and Fellows of Harvard College

The Investor Responsibility Research Center Institute (IRRCi) is a nonprofit research organization that funds academic and practitioner research that enables investors, policymakers, and other stakeholders to make data-driven decisions. IRRCi research covers a wide range of topics of interest to investors, is objective, unbiased, and disseminated widely.

The Labor and Worklife Program (LWP) is Harvard University’s forum for research and teaching on the world of work and its implications for society. Located at the Harvard Law School, the LWP brings together scholars and policy experts from a variety of disciplines to analyze critical labor issues in the law, economy, and society. The LWP also provides unique education for labor leaders throughout the world via the oldest executive training program at Harvard University, the Harvard Trade Union Program, founded in 1942. As a multidisciplinary research and policy network, the LWP organizes projects and programs that seek to understand critical changes in labor markets and labor law, and to analyze the role of unions, business, and government as they affect the world of work. By engaging scholars, students, and members of the labor community, the program coordinates legal, educational, and cultural activities designed to improve the quality of work life.

For more information, please contact:

Jon LukomnikExecutive Director, Investor Responsibility Research Center Institute40 Wall Street, 28th Floor | New York, New York, 10005(+1) [email protected] | www.irrcinstitute.org | @IRRCResearch

Larry Beeferman and Aaron BernsteinLabor and Worklife Program, Harvard Law School125 Mt. Auburn Street, 3rd Floor | Cambridge, MA 02138(+1) 617.495.9265 | [email protected]; [email protected] | http://www.law.harvard.edu/programs/lwp/

Larry Beeferman, report co-author, has served as director of the Pensions and Capital Stewardship Project since its inception in 2004 at the Labor and Worklife Program (LWP) at Harvard Law School. The Project, through conferences, trustee training, and research and publications educates and informs workers, scholars, researchers, and practitioners on issues of retirement security, including employment-based retirement plans, and of pension fund governance, management, investment and related matters. As director, Beeferman has written or co-authored papers on infrastructure investment, labor and human rights and investor decision-making, private equity and labor, labor-friendly pension fund investments, capital stewardship and labor voice, the fiduciary duty of pension fund trustees, the Dodd-Frank financial reform legislation, automatic enrollment in U.S. defined contribution plans and recent changes to Brazilian public sector pension plans. He has given many presentations and taught classes on these and related topics. His career prior to coming to the LWP, included service as head of the Asset Development Institute at the Heller School for Social Policy and Management at Brandeis University, Professor of Law at the Massachusetts School of Law, and Associate Counsel to the Special Commission Concerning State and Country Buildings. He was awarded a J.D. from Harvard Law School and a Ph.D. in Applied Physics from Harvard University. He has written books, papers, and articles on law, social policy, and other matters.

Aaron Bernstein, senior research fellow with the Labor and Worklife Program at Harvard Law School co-authored this study. He co-founded the Pensions and Capital Stewardship Project’s Investor Initiative in 2008 to help pension funds and other investors analyze the long-term investment risks of social factors such as labor and human rights and human capital. He has written several studies available on the Project’s publications page, and he speaks on the subject at conferences, seminars and workshops. He is also the editor of Global Proxy Watch, a corporate governance newsletter for institutional investors. Bernstein left BusinessWeek magazine in 2006 after a 23-year career as an editor and senior writer covering workplace and social issues. He received numerous journalism awards, including the Overseas Press Club, the Gerald Loeb, the George Polk, the New York Press Club, and the Sidney Hillman. Before joining BusinessWeek, he worked at Forbes and for United Press in London. He received a BA in Politics and Economics from the University of California at Santa Cruz and did graduate work in Political and Legal Theory for two years at Oxford University. He is the author of a book entitled “Grounded: Frank Lorenzo and the Destruction of Eastern Airlines,” and the co-author of “In the Company of Owners: The Truth About Stock Options.” Bernstein was a Wertheim fellow for the LWP during 2007-2008.

About the Authors

The Materiality of Human Capital to Corporate Financial Performance 2

Institutional investors have become increasingly interested in analyzing long-term investment risks and rewards posed by environmental, social and governance (ESG) factors. A growing body of data and analytical tools has been developed to assist in the task, but the focus has largely been on environmental and governance matters. This paper helps fill in the gap on social factors, specifically those involving how companies manage workplace relationships, a topic often referred to broadly as human capital or human resource (HR) management.1 We examine both a wide range of HR policies and separately those that relate directly to employee training.2

Our survey of the literature on human capital found 92 empirical studies that examined the relationship between HR polices and financial outcomes such as return on equity, return on investment and profit margins. We conclude that there is sufficient evidence of human capital materiality to financial performance to warrant inclusion in standard investment analysis. However, we also find that doing so remains a challenge for a number of reasons. These range from the fact that companies do not provide investors with comparable data to a lack of consensus over which combinations of policies have the most impact on financial outcomes.

This paper is organized as follows. The introduction discusses why investors seek data on social factors and examines the conceptual and methodological problems with which researchers have wrestled in analyzing training and human resource management systems more generally. Section One reviews the literature on training and explains why this subject has been treated as a distinct topic separate from those concerned with other HR policies. Section Two reviews research on the latter. Section Three considers some of the challenges investors face in attempting to apply human capital metrics to investment analysis and offers suggestions about the kind of quantitative data and other information investors might want to seek from companies. The Conclusion summarizes our findings that corporate training and other HR policies, if implemented correctly, can enhance financial performance. Investors who seek to maximize the impact of their integration of ESG factors into corporate analysis ought to consider these financially relevant factors.

Human Capital Materiality

IntroductionWhy HR Policies are Important to InvestorsHuman capital management has become widely accepted as a key component of corporate strategy. Executives, management consultants and governments have embraced the importance of corporate HR policies, including employee training. The topic has been the focus of extensive research as well. Hundreds of academic and practitioner studies undertaken in dozens of countries have examined the operational and financial benefits to companies that adopt various kinds of HR policies.

But this perspective has not carried over to the investment community in any systematic fashion. It does not engage in critical evaluation of HR management as a standard element of investment analysis. Nor have investors pressed companies to report publicly on workplace-related policies and outcomes as they have on other ESG topics such as the environment and corporate governance.

There are multiple reasons why this is the case. HR management is a complex issue that can vary by company, industry and country. There remains considerable debate about which HR strategies are the most effective in particular contexts. There is a paucity of publicly available data that would allow meaningful analysis of companies or comparisons among them. The data that does exist is typically not audited or otherwise subject to external assurance.

Underlying these concerns is the possibility that most institutional investors are largely unaware of the extensive evidence that already exists about the materiality of human capital factors. This is understandable given that most of the studies in the field have not been framed from the perspective of investment analysis. Some researchers have included privately held companies in their analysis along with publicly traded ones, which precludes or makes difficult analysis of standard investment outcomes. The challenge is compounded by the way companies often roll out HR policies in only some work sites or units, or only for certain classes of employees, leading many studies to focus on divisions of companies and even individual offices or factories.

Human capital research has been undertaken in hundreds of studies encompassing a multitude of disciplines: Numerous studies have been done in the fields of economics, labor studies, human resource management, psychology and sociology, but investment outcomes have been a concern only in a minority of them. For example, a 2010 literature review of 66 papers on training done between 1991 and 2007 identified outcomes that included productivity, sales growth, employee commitment, value added per worker,



firm present value, turnover, market share, export sales growth, customer satisfaction, sales per employee, employee satisfaction, client satisfaction, owner/shareholder satisfaction, absenteeism, product and services quality, work performance, cooperation, discipline, new product development and equipment downtime.3 A 2013 review of research on the relationship between HR policy and these kinds of firm operational performance found 248 articles assessing an equally wide range of outcomes.4 These counts exclude the related but largely separate field of employee job satisfaction, which a 2009 paper estimated had been the subject of 10,000 studies and articles.5

Our review sorted through this field to highlight research that included traditional corporate financial performance indicators widely used by institutional investors. We identified 92 studies that assess one or more of these investment outcomes, 36 specifically on training and 56 on HR systems more generally. The financial metrics include total shareholder return, return on assets, return on earnings, return on investment, return on capital employed, profitability and Tobin’s Q. We have excluded productivity, even though it was the most common result assessed in this body of research. Many economists consider output per worker to be the most rigorous and reliable way to assess corporate performance across firm samples. Indeed, numerous studies treat productivity and financial performance as synonymous. We nonetheless chose to omit productivity outcomes because it is not a standard variable used in commercial investment analysis. Even so, the 92 papers that did focus on standard investment metrics still offer a good sense of the strengths and weaknesses of the findings in this field.

We restricted our review to studies that used conventional investment indicators to emphasize the conclusion that human capital is material under definitions acceptable to the U.S. Securities and Exchange Commission and U.S. securities law.6 Investor interest in non-financial risks and rewards has precipitated a proliferation of efforts to develop the field. These considerations are often characterized as environmental, social and governance (ESG) factors, and human capital typically is seen as falling into the social category. The scope of ESG factors and the rationales for taking account of them remain under development. Some investors employ the term to describe factors partly or even largely in normative terms while others define it as focused strictly on elements that are material to corporate financial outcomes.

The majority of these 92 studies found positive correlations between training and HR policies with investment outcomes (Table 1). We discovered just one with only negative findings. Seven others found no correlations and another seventeen uncovered a mix of positive outcomes and either no correlations or negative ones.

Topic Financial Effect

Positive Mixed None Negative

Training (36 Studies) 22 8 5 1

HR Policy (56 Studies) 45 9 2 0

Total Number of Studies 67 17 7 1

The simple count of studies presented here does not take into consideration the quality of the studies’ research methodology or the robustness of their findings. Many studies also present multiple findings; when possible we focus on those the authors present as their primary ones.

Table 1: Human Capital Studies

Although the research varies in depth and quality, in aggregate the literature offers considerable empirical evidence that human capital policies can be material to corporate performance. The total number of positive findings is given added weight by the diversity of industries and countries to which they apply.7 The results suggest to investors who may not have been persuaded by or aware of long-standing assertions about human capital materiality that they ought to reexamine the data.

One of the most forceful statements of this conclusion came in a 2003 report by a Task Force on Human Capital Management (HCM) established by the British Secretary of State for Trade and Industry, which included several high-level executives of prominent British companies. It concluded:

“HCM should not be regarded solely as an internal matter for management. For most organisations the link between HCM policies and practices and performance is sufficiently central to be a material factor whose disclosure might reasonably be expected to influence assessments of their value and effective stewardship by management. In such cases disclosure increases the value of financial reports and will be important for the effective operation of capital markets.”8

The Global Reporting Initiative (GRI), the largest and most widely used ESG reporting entity, has encompassed both the “value” (financial impact) and “values” (normative behavior) perspectives, reflecting its status as a multi-stakeholder association of advocacy groups, nonprofits, investors and others. In 2013 the GRI released a fourth-generation version of its guidelines that put more emphasis on investment materiality while retaining the normative principles which have guided the group since its founding by two US nonprofits in 1997.9



Even entities whose goal is to establish ESG materiality standards do not always eschew normative considerations. The Sustainability Accounting Standards Board (SASB) was founded in 2011 with just such a mission.10 However it has not relied primarily on conventional academic research to achieve this goal. Instead SASB has combined keyword evidence searches with a crowd-sourcing effort that draws on input from a wide variety of expert volunteers, including socially responsible investment funds, activist groups and labor unions whose perspective often includes normative concerns, as well as investors and corporate participants who might be assumed to have more concern with financial impacts.11 The overlap of normative and financial perspectives is also evident in groups such as the United Nations-sponsored Principles for Responsible Investment initiative, which states that ESG issues can not only ”affect the performance of investment portfolios” but also “may better align investors with broader objectives of society.”12

MaterialityOur decision to focus only on traditional investment outcomes is intended to address a common misunderstanding that the materiality of ESG factors in general, and human capital ones in particular, is not yet backed up by research pertinent to mainstream investors. Although there is sparse evidence of such materiality for numerous social factors, the papers reviewed here offer substantiation of the correlation to financial outcomes for training and HR policies more generally.

A wider appreciation of this literature can help the investment industry assess priorities for corporate ESG reporting. ESG data providers offer investors information about corporate workforce policies, but the choice of variables often appears to be a function not of materiality but of what companies decide to make available or are required to report by regulators. One prominent example is an annual report first issued in 2012 by Corporate Knights Capital, an investment advisory and research firm based in Toronto. Its third report published in 2014, Measuring Sustainability Disclosure: Ranking the World’s Stock Exchanges, used Bloomberg data to assess information provided by 4,609 large companies listed on 46 exchanges. The document focused on seven so-called “first-generation” sustainability indicators, which it selected “because they are objective measures of corporate sustainability performance that are broadly relevant for companies in all industries.”13 Three of the indicators concern workforce issues: employee turnover, injury rates and corporate payroll.14 The report, which was backed by Aviva, a large British insurer that has been actively involved in promoting better ESG reporting, as well as by Standard and Poor’s and the Association of Chartered Certified Accountants, urged stock exchanges and “policy-makers of all description”

to “encourage or mandate listed companies (and large listed companies in particular) to measure and publicly disclose their performance on the seven first-generation sustainability indicators.”15

Yet Corporate Knights offers no argument that these seven indicators are material to investors or even how precisely they were selected in the first place. The report pointed to no research on the materiality of any indicators – despite the decades of research we discuss below demonstrating the importance of training and other HR management policies to financial outcomes.

The question of how to define and prioritize material social factor indicators is of growing importance in light of regulatory efforts to mandate corporate ESG reporting. In 2014 the European Union passed legislation requiring companies with 500 or more employees to disclose ESG information or explain why they do not. It instructs the European Commission to develop guidance on which indicators companies should use and consult stakeholders during the process, which must be completed before the rules take effect in 2017.16 Investors now have an immediate interest in advising the Commission on which indicators should be used. It would seem logical, at a minimum, to consider asking companies to report on indicators found to be material based on extensive research, or explain why they do not.

The same may hold true with a broader effort involving a petition asking stock exchanges around the world to adopt ESG disclosure listing requirements. It was developed by the Investor Network on Climate Risk and submitted for comment in 2014 to the World Federation of Exchanges, an association of 64 exchanges.17 The WFE set up a Sustainability Working Group that year in part to consider the proposal.18 One potential outcome of the petition is a recommendation to exchanges by the International Organization of Securities Commissions, an association of 120 regulators around the world that sets standards for the securities industry.19

Links to PerformanceInvestigations of links between corporate performance and training and HR management date to at least the late 1930s, when researchers looked mostly at outcomes of company policies such as employee job satisfaction but failed to find much.20 Interest picked up again in the early 1960s after Nobel Laureate economist Theodore Schultz began using the term human capital to describe the investments and systems companies used to train and manage employees.21 Then in 1964 another Laureate economist, Gary Becker, wrote a seminal book entitled Human Capital.22 Many of the papers that followed his



work looked at productivity. Researchers began to examine the investment outcomes covered in this review in the 1980s.

Most of the initial research sought to find correlations between measures of performance and a wide spectrum of discrete HR policies regarding employee training, team systems, profit-sharing, employee ownership or hiring, retention and promotion. While many discovered positive effects, some did not. By the 1990s there was a gathering consensus that adopting just one type of policy often might not deliver value, or might not produce maximum value. Instead, the emerging view was that companies derived the most benefit from bundling groups of policies together in a synergistic approach. Many researchers adopted the term “high performance work system” to characterize certain bundles that typically include elements such as teams, worker participation and some kind of profit- or gain-sharing.

In 1995 a study by Rutgers University academic Mark Huselid launched a subgenre of research focused on the links between high performance work systems and financial performance.23 Examining 968 publicly traded U.S. firms with 100 or more employees, he found positive correlations to both Tobin’s Q and return on capital employed. Since then hundreds of studies have looked at such links, although many included productivity in their definition of firm performance.24

Some researchers continue to use the phrase high performance work systems while others define bundles of policies more broadly and may or may not include those labeled as high performance. Either way, the idea that HR policies are the most effective when used together has since evolved into the standard perspective adopted by most recent research. The review of the field in this paper includes both studies that examine the materiality of discrete HR policies as well as those that assess bundles of policies.

We have broken out training as a separate category because much of the research has done the same. Dozens of studies have focused on training as a stand-alone policy unconnected to a company’s other HR practices. While many bundle analyses include training as one of the elements, the evidence is substantial that even considered in isolation, training is frequently associated with higher profits for firms. As a result, the first section of this paper deals with that body of research while the second addresses other HR policies considered individually and collectively. Studies that include training as part of a bundle are covered in the second section.

There are several other workforce-related topics that have been extensively examined for links to corporate performance. In addition to health and safety this includes

diversity of both employees and of boards of directors. We have excluded them to keep the project to a manageable scope.

We have included studies from dozens of countries, on both training and broader HR approaches. More than half of those we selected were done on companies in the US and the UK, where research on these subjects has been more extensive. (This also may be at least in part a function of the fact that we restricted our search to papers written in English.)

A few general notes of caution apply to most of the literature reviewed. There has been limited effort to link the vast body of work on employee views and job satisfaction to investment performance, despite recent suggestions that this is an important complement to surveys of corporate management views on HR policies.25 In addition, authors often have had far more to say about the correlation of HR policies and financial performance and rather less about the size of the effect. And while many such correlations have been found, the question of causality remains a topic of debate, with some researchers suggesting that better training and HR policies may follow from superior financial performance rather than be caused by it. Another is whether such policies improve financial performance directly, or only if they are adopted in conjunction with other steps such as a superior business or corporate responsibility strategy. There also has been limited attention to whether there might be diminishing returns to increasing the scale of particular HR policies.26 There are as well a variety of questions about the quality of the data and of the methodologies employed. We cite the concerns that are appropriate to each of the studies as we review them.



Section One: TrainingGary Becker’s 1964 book, Human Capital, was among the first attempts to argue that company-paid training is not just an expense but is also an investment akin to other capital costs.27 Economists and other experts have been trying ever since to sort out exactly how that works. Before his book the prevailing view was that formal or informal on-the-job training was analogous to any other form of education such as high school or college. The contention was that employees reaped most of the benefit from this education, since they could always leave for a higher-paying position at another company after the training enhanced their skills. So even if employers actually provided the training, they offset the cost by paying lower wages until the training was complete. Becker altered this perspective by arguing that companies and employees both benefit from training, even if the employee enjoys subsequent wage gains as a result.28

Our review of the research on training is focused on the benefits it can bring to companies, specifically as they affect financial performance. We have identified 36 studies that analyzed links between training and investment outcomes. Five found no correlations and eight described a mix of positive associations, negative ones, and no correlations. One reported only negative correlations. The remaining 22 concluded that training is associated only with superior investment outcomes. These findings have come in multiple countries and industries and in studies stretching back more than three decades.

Although the evidence is strongly suggestive of a payoff to companies, researchers continue to debate exactly how it occurs. The predominant theory holds that training enhances employee knowledge, skills, and abilities, which improves outcomes such as productivity, product and service quality, and customer satisfaction.29 These in turn can lead to higher sales, profitability and ultimately stock valuations. Most of the studies employ regression analysis to associate training with these outcomes. Regressions may explore not only the direct relationship between training and other measures of financial performance but also that relationship contingent on other factors such as the characteristics of the firms and the presence or absence of other HR policies. It has proven more difficult to weigh the benefits against the many costs of training, such as materials, trainers, lost work time and added managerial expense. Moreover, as reflected in Tables 2 and 3, researchers have used several different measures to define what is meant by training.

The challenge has been compounded by the general nature of much training, which enhances employees knowledge and skills that can be employed in many contexts as


How Training is MeasuredAlthough companies are not typically required to publicly disclose their training expenditures (let alone training-related policies or how those actually are implemented), many are willing to do so when asked (sometimes subject to confidentiality requirements). The majority of the studies reviewed in this paper obtained training data from surveys and questionnaires sent to companies. Typically, the respondents were corporate HR managers, although sometimes other managers or executives responded. While the response rates varied, thousands of public and private companies in dozens of countries have answered the requests over several decades.

opposed to firm-specific training, which teaches them how to carry out particular tasks at an individual employer. Numerous studies have found a majority of training to be general in nature, perhaps as much as 60% to 70%.30 Companies typically bear the cost, even though employees remain free to employ their newfound skills at other companies. More specific skills training is, of course, less transferable to new employment.

These complex relationships have made it difficult to undertake cost-benefit analyses of training. Part of the problem stems from corporate accounting, which treats training as an expense rather than an investment. Companies are not required to report training expenditures as a discrete item, so it often is lumped in with other overhead. This is an understandable approach since the payoff to training is difficult for an individual company to quantify. If companies could treat any and all training expenditures as an asset, it could lead to overstated book value, overstated earnings and excessive dividends and management bonuses.31 Nonetheless, the accounting treatment has hampered efforts to assess the economic returns to companies. As one study concluded: “In summary, economists view investment expenditures as any outlay made by managers in the expectation of future benefits, whereas GAAP (Generally Accepted Accounting Principles) rules determine an investment by reference to an internal rule set, giving rise to a potential disconnect between what firms do and what GAAP reports.”32

The lack of consistent reporting has led to challenges in linking training to corporate financial performance. “Unlike all other major categories of investments that firms make to enhance their future productivity and profitability (e.g., physical plant and equipment, research and development), investments in developing human capital are neither separately accounted for, nor are they publicly reported,” one study observed. “These investments are thus essentially invisible to most investors (with the important exception of the fact that they raise costs in some indeterminate way).”33


The surveys have allowed researchers to observe links between training and financial performance that can be obscure at an individual company. The typical approach has been to use standard regression analysis to compare training among groups of companies. Most of the 36 studies we selected for review measured training expenditures, either per employee, per firm, or as a percent of each company’s total payroll. Some used the existence of a training policy, the fraction of the workforce trained, or time devoted to training. There also was variation among types of training. Some research focused on formal skill instruction while other papers looked at general on-the-job training. Training recipients differed as well, from entry-level employees to specific groups such as bank tellers or production workers. Although some of the studies found no relation to financial outcomes, there is no indication that a correlation or lack thereof was a result of the training variables employed.34

Another complication is that many studies do not measure financial performance using the audited public data. One reason is that many include private corporations that do not release such information. Another is that some researchers have studied divisions of companies and even individual factories and offices. Because detailed financial results are usually unavailable for such entities, analysts have had to rely on survey questions answered by company officials. As a result, most of the studies use perceived measures of profitability as reported by company executives in the survey. Often the questions are asked in different ways, such as profits over the past year or in relation to competitors.

Although perceptual surveys are less desirable than reported financial reports for investment analysis, it is not clear how much they undermine the findings that training is material to companies’ financial performance. On the one hand, some experts have found that estimates for the link to performance are higher from surveys, suggesting that managers may overestimate the link to policies such as training.35 On the other hand, however, other researchers have conducted separate studies comparing self-reported financial results with publicly reported ones and found them to be largely consistent, according to a 2008 paper.36 The authors went on to argue that publicly reported results themselves can distort comparisons among firms, particularly in cross-national studies involving countries with different reporting and accounting requirements. Although the issue is not completely settled, surveys are widely used in academic research as a way to assess many aspects of corporate performance, not just training and human capital.

The measures of financial performance on which researchers have focused have varied. Only a handful of training studies have used stock performance. Four of the seven we identified were co-authored by Laurie Bassi; mostly using data from annual surveys by the American Society for Training and Development (ASTD), where she was Director


of Research in the late 1990s.37 For example, a 2004 study she co-authored called The Impact of U.S. Firms’ Investments in Human Capital on Stock Prices examined ASTD data on training by 388 companies between 1996 and 1998. It found that those in the top quartile of formal employee education and training expenditures averaged annual stock price returns between 1996 and 1998 of 31% while those in the bottom quartile averaged 15%.38

Study Sample Training Indicator

Financial Indicator

Results Country

American Bankers. 2004.

Survey of 17 banks

Training ex-penditure per FTE

Profitability, TSR, ROA, ROI

Banks in top 50% of training expen-diture per FTE performed better on all indicators

US

Aragon-Sanchez. 2003.

Survey of 457 companies with 10 to 250 employees

Training policy Profitability Positive Finland, Netherlands, Portugal, Spain and UK

Aragon. 2013. Survey of 316 large firms

Training policy ROA Mixed Spain

Bartel. 1995. Analyses of personnel re-cords of 19,000 employees at a large US manu-facturing firm

Training expenditure per employee

ROI* Positive US

Bartel. 2000. Corporate records of a New Jersey manufacturing firm and a New Jersey service firm and of the Garrett Engine unit of Allied Signal


ROI* Positive US

Bassi. 2001. Survey of 575 publicly traded firms

Trainingexpenditure per employee

Stock price A $100 increase in training expenditure per employee in-creased the annual stock price by 0.8 percentage points

US

Bassi. 2002. Survey of 575 public companies


TSR Firms in the top half of training expendi-ture per employee in one year had a mean TSR the fol-lowing year of 37%, vs. 20% for those in the bottom half

US

Table 2: Training Studies



Financial Indicator

Results Country

Bassi. 2004. Survey of 388 companies


Stock price, ROA

Firms in the top quartile of training expenditures per employee had annual stock price returns between 1996 and 1998 of 31% while those in the bottom quartile had 15%. Return on assets averaged 5.3% for the top quartile and 4.2% for the bottom.

US

Bassi. 2009. Surveys of 30 banks


Stock price Training expendi-ture per employee in one year corre-lates to 21% of the variation in stock price performance in the following year relative to competitors

US

Bernthal. 2006.

Survey of 127 firms

Training policy ROA, ROE, profitability

Firms with high-quality leadership training programs and manage-ment succession program performed better on all indica-tors

US, Canada

Blandy. 2001. Survey of 41 firms

Training quantity and quality

Profitability Positive Australia

Bosworth. 2002.

Longitudinal surveys of 3,569 busi-ness units with less than 200 employees between 1994 and 1998

Training expenditure

Profitability Positive Australia

Bourne. 2008. Survey of 196 firms

Training policy Profitability, ROA

No Correlation UK

Chen. 2008. Data on 802 public account-ing firms from 1992 to 1995

Training hours per employee

Profitability Positive Taiwan

Chochard. 2011.

Interviews with supervisors and participants of leadership train-ing programs at ten Swiss firms


ROI Positive Switzerland

Cosh. 2003. Survey of 2,500 firms

Training expenditure per firm and per employee

Profitability Mixed UK



Financial Indicator

Results Country

d’Archimoles. 1997.

Survey of 42 firms

Training expenditure as a percent of total wage costs

ROCE Positive France

Danvila del Valle. 2009.

Survey of 40 private security firms

Training expenditure per employee, number of dif-ferent training courses given, training hours per year per employee

Profitability Positive Spain

Doucouliagos. 2000.

Case studies of five large and two small em-ployers collec-tively employing 70,000 people

Training expenditure per firm

ROI Positive Australia

Faems. 2005. Survey of 416 firms with 10 to 100 employees

Training policy ROE No Correlation Belgium

Hansson. 2007.

Survey of 5,824 firms with 200 or more employees

Training expenditure as a percent of wage bill and as percent of workforce trained

Profitability Mixed 26 countries

Jones. 2011. Data from 233 banks

Training expenditure per employee; training days per employee

Profitability No Correlation Finland

Kim. 2013. Survey of 359 firms with more than 100 em-ployees

Training policy Profitability Mixed South Korea

Leitner. 2001. Survey of 100 firms employing 20 to 500 people

Training policy Profitability Positive Austria

Meschi. 1998 Survey of 102 firms with 250 or more employees

Training expenditure as a percent of wage bill; training policy

ROI No Correlation France

Mohrenweiser. 2009.

Government data for 1,879 private-sector firms with 20 or more employees

Apprentices as a share of semi- and un-skilled workforce

Profitability Mixed Germany



Financial Indicator

Results Country

Morrow. 1997. Evaluations of 18 mana-gerial, sales and technical employee train-ing programs conducted over four years by a Fortune 500 firm


ROI Positive US

Newkirk-Moore. 1998.

Survey of 152 community banks with less than US$500 million in assets

Training policy ROA, ROE Mixed US

Park. 2011. Survey of 454 firms with more than 100 employees in 2005 and 2007


Profitability, ROA

Positive South Korea

Percival. 2013.

Annual surveys of 3,528 firms from 1999 to 2005


ROI* Mixed Canada

Storey. 2002. Survey of 314 firms with sales between £6 million and £500 million

Training policy ROCE Positive UK

Sung. 2014. Surveys of managers and employees at 207 manufacturers

Training policy ROA Mixed South Korea

Úbeda-Garcia. 2013.

Survey of 112 hotels

Training policy Profitability Positive Spain

Vanhala. 2006 Surveys of 91 firms in the metal industry and retail trade between 1997 and 2000

Training policy Profitability Positive Finland

Wright. 1999. Survey of 38 refineries

Training policy Profitability, sales growth

Negative US

Zwick. 2007. Government survey data from 1997 to 2004 covering up to 6,000 firms

Apprentices as a share of workforce

Profitability No Correlation Germany

ROA=return on assetsROI=return on investmentROCE=return on capital employedFTE=full-time employee*ROI here refers to the company’s returns on its training expenditures


The studies reviewed used a variety of measures of profitability, including return on assets, return on investment, return on equity and profit margins. The majority found positive correlations to training. One of the most comprehensive was a study of Australian companies that examined training expenditure per firm and profitability at 3,569 firms with fewer than 200 employees. It had a large sample size as well as access to data between 1994 and 1998, allowing it to track unit performance across time instead of just taking a one-time snapshot as most studies do (i.e., it was a longitudinal study instead of a cross-sectional one). It found that firms which had increased training in one year reported significantly higher profitability the following year.39

The study with the largest sample size also found positive links between training and profits, although the robustness of its sample was offset by its cross-sectional nature. It used data from the Cranet survey, which was established in 1989 by the UK, Germany, France, Sweden and Spain and coordinated by the Centre for European Human Resource Management at the Cranfield School of Management in Cranfield, England. Although the survey has been conducted multiple times, each one is based on a random sample of companies and therefore does not allow for longitudinal analysis. The 2007 study we reviewed used 1999 data on 5,824 private-sector companies with 200 or more employees in 26 countries, most of them European with a few others, including Australia, Israel, Japan and Tunisia (but not the United States). It found positive correlations between the percent of total wages spent on training and whether firms described themselves as in the top 10% of their industry in profits. The study also looked at a variety of other factors that might correlate to self-described profit performance. It concluded: “Apart from the firm’s past profitability, the amount invested in training is the most important factor in explaining the probability of belonging to the top 10 per cent in profitability in an industry. This result also contributes to the existing literature by confirming previous country-based findings on the profitability of training investments that, from a global sample of firms, suggest that training investments generate considerable gains.”40

Another large-sample cross-sectional study found positive correlations between training expenditure per firm and profit margins over the prior three years, although it uncovered none of statistical significance for expenditures per employee.41 This research also was significant because it assessed training in Britain, which has one of the most comprehensive national training efforts. The government began a program in 1990 called Investors in People (IIP) that was intended to create “a national framework to link the process of setting business objectives with staff development to improve business performance.”42 IIP has changed over the years but essentially bestows public recognition through accreditation on firms that engage in training and other employee development efforts.43 By 1999, the time of the survey used in the study, 29% of the 2,500 firms


sampled already participated in IIP. The study found that IIP participation was closely associated with effective training programs, and that firms spending more on training had higher profit margins. A more recent survey in 2010 of studies of IIP largely confirmed the positive correlations between training expenditures and financial performance, including three that found links to publicly reported profits.44

Five studies found no correlations, although most looked at training as part of an examination of other HR policies which did turn out to have positive associations with investment outcomes. This was the case with one of the largest, a 2007 study that used German government survey data from 1997 to 2003 ranging from 9,000 business establishments in 1997 to 16,000 in 2004. It found no correlation between profits and full-time apprentices as a share of firm’s workforce.45 The paper did find positive associations with the existence of works councils.46

Eight papers showed a mix of outcomes. The largest involved 2009 research that used the same German government data set, this time looking at 1,879 firms with 20 or more employees. It measured training by looking at companies’ use of apprentices as a share of employment of semi- and un-skilled workers. This ratio turned up positive associations with gross profits per capita in trade, commercial and construction occupations, but negative correlations for manufacturing occupations.47

We found one study that came up with negative correlations between training and investment outcomes. It examined a group of seven training policies, such as hours of training and amount of money spent on training, and found that they were associated with lower profit margins in 1993 as well as lower five-year profit growth ending in that year. However, this research, which involved a survey of 38 US refineries, also found positive results for other HR policies unconnected to training such as pay for performance, appraisals and employee participation systems. The authors were surprised by the training results and speculated that they may have stemmed from the capital-intensive nature of refining. “In many cases the technology used is aimed at decreasing the skill requirements and discretion of operators,” they said.48

Corporate StrategyDespite the accumulation of positive findings over many markets and years, treating training as a factor in investment decisions still presents a variety of challenges. Because the majority of research on the topic is written from the perspective of corporate managers, much of it is concerned with performance metrics that are not easily quantified or publicly reportable in a standardized fashion. (There are other issues as to whether


such managers are the appropriate persons to be asked about policies and the accuracy of their answers.49) Although most of the studies we reviewed found superior investment outcomes among firms that train more or have well-developed training policies, such factors are only starting points. While expenditures offer a robust signal investors can use to rate corporate behavior, other factors come into play as well. Among them: how a company fits training into its competitive strategy, whether it can make effective use of the employees it trains and whether they factor in the national training and education policies of the countries in which they operate.

Several researchers have made this point, arguing, for example, that companies should train only as part of a corporate strategy based on higher skills. “The implication is that the company must first develop a business strategy in which the skills of its employees are seen as providing a source of competitive advantage,” suggested a 2006 paper on training and strategy.

“Our model therefore suggests that it is not always useful to exhort all employers to train more. For some employers (with their specific competitive strategy), training beyond the operational level is pointless and counter-productive. Resources devoted to such an ‘undifferentiated’ skills policy are likely to be wasteful. Perhaps a first step here is to determine how these competitive strategies and their component technical and interpersonal relations differ between sectors. If, as some evidence suggests, business strategies vary significantly across sectors, then there will be little point in spending resources on convincing employers of the need for training if their business strategies are centred around standardised technical relations and task focussed interpersonal relations.”50

Others have expanded on this point, by, for instance, suggesting that companies can reap the most benefit if they develop formal plans to align training with their strategic needs.51

Others suggest that companies should tailor training to different markets in which they operate. Another study using the 1999 Cranet data described above found no correlation between training and self-reported profitability relative to other firms. This one, done in 2008, focused on a sample of 5,189 businesses in 14 European countries. It looked at training policy rather than expenditures and used a different definition of profitability than the first Cranet study, but found no connection. Instead, the authors determined that national policies affected training results – even at a very broad level such as the share of GDP spent on education. Their findings: “in countries that spend more on education, employee training has a negative effect on firm performance, while in countries that spend less on education, employee training has a positive effect on firm performance.” Companies therefore may


waste resources training employees in countries with good education systems but profit from doing so in markets with significant skills gaps. “So, if firms want to increase their performance, they need to take into account the national levels of expenditure on education and align them with organizational-level training,” they wrote.52

The type of training can also make a difference in terms of its impact, and it is not always easy for investors or others outside the company to understand the nature of the training. A counter-intuitive argument made by some analysts holds that companies gain more if they provide general rather than firm-specific training. Although we found no comparisons of the two kinds of training relative to investment outcomes, a 1999 paper on surveys of 215 nationally representative Irish companies found positive effects of general training on productivity growth but no effects from specific training. The conclusion they draw is that employees put less effort into specific training because it does not benefit them as much.

“As we argued above, employees are not mechanical black-boxes into whom training is injected. Rather they are rational players who must choose the amount of energy they will devote to turning the training they receive into additions to their human capital. Training which increases an individual’s wage with both the existing employer and potential employers provides greater incentives for effort than training which only increases wages with the existing employer. This view of the training process is true whether the employees pay for the training themselves, as predicted by Becker, or the employer pays.”53

Their argument may help to explain why so much of employer-provided training is general in nature and thus of benefit to future employers as well as the current employer, and perhaps as well as to the economy as a whole.54

The general nature of much training raises another question researchers have grappled with for years, namely how much of what employees learn in training is actually applied in practice. The issue has spurred decades of research into what is often referred to as “the transfer of training,” meaning how employees transfer the training they receive to the job they perform. As early as the 1980s academics found evidence that significant amounts of training were wasted because companies paid too little attention to the transfer challenge. This was explained in a 2014 study that brought the problem into focus from the perspective of firm performance (although it did not use investment outcomes as the yardstick). It surveyed 150 professionals who belonged to a national training association in Canada and found that those who described higher degrees of transfer also reported that their firms performed better than rivals over the prior three years on outcomes such as quality of products and services and customer satisfaction.55


The transfer issue poses a challenge for investors, who are likely to find it difficult to obtain reliable information on how companies deal with it. Information on training expenditures and even policy is self-reported and unaudited for the most part, but it involves hard numbers and written documents. Reporting on transfer success requires firms to make many more judgment calls that seem likely to inhibit vigorous reporting.

Cost-Benefit AnalysisEven if companies successfully transfer training and reap financial gain from it, there remains a question about its cost. In theory this should be answered by the accumulation of studies finding higher profits at firms that train more. The regression analyses most employ are designed to eliminate other potential factors that might lift returns at such companies, which implies that the better returns are net of costs. But this is not assessed directly.

Over the years, several studies addressed the cost/benefit issue by attempting to assess all the costs companies incur when they train and to capture the benefits that can result. One of the earliest studies to focus on a specific company was done with the personnel records of 19,000 employees at a large U.S. manufacturing firm.56 The author used the documents

“to calculate the per-participant direct costs of a day of training, which includes the salaries of the trainers and the costs of materials, room, and board. The indirect costs of training were calculated from data on the salaries of the trainees. Direct and indirect costs were then summed to determine the per-participant training cost. On average, during the 1986-90 time period, it cost the company $1,440 to provide 1 day of training to an employee.”

The author then determined the wage gains attributed to the training, which she argued gave a conservative estimate of the productivity improvements the firm saw from the training. She calculated different estimates of the company’s net return using a variety of assumptions about how quickly the skills employees acquired depreciated over time. For example, a depreciation rate of 10% a year implied that the company earned “34.6% on employee development training and 36.6% on technical training.” The writer came up with similar results in a subsequent study that analyzed corporate records of a New Jersey manufacturing firm, a New Jersey service firm and the Garrett Engine unit of Allied Signal.57

Three other studies came to similar positive conclusions about the net returns to corporate training expenditures. One looked at leadership training programs at ten Swiss firms.58 Another analyzed seven Australian firms that collectively employed


70,000 people.59 A third evaluated eighteen managerial, sales and technical employee training programs conducted over four years by a Fortune 500 firm.60

Although these findings bolster those who found higher profits at companies that trained more, they only analyze a handful of firms. This leaves open the possibility that some or even a significant portion of companies might lose more than they gain from training programs. Some research has found that to be the case. A 1996 study attempted to quantify training costs and benefits through a survey of 50 Canadian organizations, 42 of them mostly small- and mid-sized companies (the other 8 were government entities). The survey asked companies to quantify benefits such as fewer injuries and absentees, lower scrap and waste rates, fewer delays, less turnover, less overtime, and increased productivity, which was quantified by assigning work hours saved. Costs included direct training and equipment cost as well as staff time to develop a training program. A benefit-to-cost formula standardized results, which showed that 30% of the organizations enjoyed extraordinary returns ranging from $10 for every $1 spent to $46.3. Another 52% saw returns of $1.1 to $9.9 per $1 spent. However, the remaining 18% were breakeven or even lost more than they gained.61

Findings such as these raise some uncertainty for outside observers such as investors, who have few avenues for determining whether corporate training programs are effective.

Still, the overall conclusion from the studies we review strongly suggests that there is a significant payoff to training. A number of researchers have argued that many companies still underinvest in their workforce skills. Bassi has maintained that at least part of the reason lies with the lack of public reporting and the resulting lack of appreciation of the importance of training by investors, as well as the time lag between when training occurs and when any benefits translate to the bottom line.

“Why would firms ignore the obvious and under-invest in this particular factor? Consider two organizations that are identical in all but one respect: Company A makes substantial investments in skills, while Company B does not. What will be evident to any investor comparing the companies’ income statements is that Company A has higher overhead (SG&A) and correspondingly lower reported earnings than Company B. What will not be evident, however, is that some of what were classified as `expenses’ for Company A is actually an investment in future productivity. Consequently, Company A’s stock prices would be expected to be lower – at least in the short-run – than Company B’s. The decision of Company A to invest in learning and development thus occurs despite pressures from financial markets. All firms – even those that have made significant human capital investments in the past – continually face this structural pressure to cut those investments in the short run to generate temporary increases in earnings.”62


Section Two: Work SystemsThe links between investment performance and how companies manage human capital overall are even more complex than the ties to employee training. The general theory is analogous: firms are more competitive if their work systems are designed well and function effectively to make the most of employee talent and skills by stimulating worker engagement and commitment on the job. But training is a relatively distinct activity, even if it can be general or firm-specific in nature. By contrast, human resource management systems consist of multiple policies and practices, sometimes dozens, which have been found to be more effective when implemented in coordination with each other as part of an HR strategy which, in turn, fits with a company’s overall competitive strategy in its industry. All this makes it more difficult to determine exactly which HR policies spur performance and in what combination. It is all the more demanding to tell whether companies with the best policies actually implement them well in practice.

The broader scope of corporate work systems has spurred many more studies than have been done on training. As mentioned in the introduction, training often is seen as just one of numerous HR policies companies should adopt. As with training, these studies have looked at many aspects of corporate performance. Forty-five of the 56 identified that focused specifically on investment outcomes uncovered positive correlations. Another nine found a mix of positive, negative or no correlations and two found no correlations. The complexity of human resource systems has introduced more uncertainty than is found in most training research. Training has typically been perceived as having the potential to more directly affect corporate profitability, while HR policies are more likely to do so through intermediate effects such as increased employee motivation, which in turn can improve productivity, customer satisfaction or the quality of goods and services a company produces. Still, the accumulated findings offer substantial empirical evidence for the materiality of HR policies. As is the case with the training literature, these findings stretch over several decades and dozens of countries, lending them added weight.


Study Sample HR Indicator Financial Indicator

Results Country

Akhtar, 2008. Survey of a 465 manufacturing and service firms

HR policy Profitability, ROI

Positive China

Bae. 2003. Survey of 680 firms

HR policy Profitability Positive Taiwan, Thailand, Korea and Singapore

Bassi. 1998. Survey of 500 firms with more than 50 employees

HPWS Profitability 88% of firms with HPWS reported higher profitability than peers vs. 60% of those without

US

Bassi. 2007. Survey of 11 financial services firms

HR policy TSR Positive US

Bauer. 2009. 2,265 bonds issued by 568 firms between 1995 and 2006

HR policy Stock volatility, credit risk

Firms with better HR management had lower stock volatility and credit risk

US

Becker. 1998. Surveys of 4,000 firms with more than 100 employees in 1992, 1994 and 1996*

HPWS ROA Positive US

Bjorkman. 2002.

Survey of 62 multinational manufacturing firms

HPWS Profitability Positive China


HPWS Profitability, ROA

Positive UK


HPWS Profitability, ROA

Positive UK

Chang. 2013. Survey of 74 manufactur-ing firms with at least 100 employees and $50 million in annual sales

HPWS ROA Positive US

Collins. 2006. Survey of 323 small firms

HR policy Profitability Firms with better HR management posted 23% higher profit growth over one year

US

Cowling. 2008.

Survey of 2,500 firms

HR policy Profitability Firms without better HR management would earn higher gross profits per employee if they adopted them

UK

Delery. 1996. Survey of 216 banks

HPWS ROA, ROE Positive US

Table 3: HR Policy Studies



Results Country

Deloitte. 2002. Data on 200 firms, 80% publicly traded

HR policy TSR Firms in top quartile of HR management averaged 15% TSR over five years; those in second and third quartiles averaged 4.98%; those in the bottom quartile averaged -10%

US, Canada

Derwall. 2007. Surveys of 633 firms in the Dow Jones Global Index

HR policy TSR, Tobin’s Q, ROA

Mixed 31 countries

Dolan. 2005. Survey of 180 large firms

HR policy ROCE Positive Spain

Ellinger. 2002. Survey of 208 manufacturing firms

HR policy ROA, ROE, Tobin’s Q

Positive US

Faems. 2005. Survey of 416 firms with 10 to 100 employees

HR policy ROE Mixed Belgium

Gooderham. 2008.

Survey of 3,821 establishments

HR policy Profitability Mixed 16 European countries

Guest. 2001. Survey of 82 firms that were members of the Involvement and Participa-tion Association

HR policy Profitability Positive UK

Guest. 2003. Survey of 366 firms with 50 or more employees

HR policy Profitability Mixed UK

Harter. 2003. Employee surveys at 44 firms

Employee engagement

Profitability Positive Australia, Canada, Hong Kong, South Korea, UK, US

Horgan. 2005. Survey of 81 Irish firms and 311 Dutch firms

HPWS Profitability Mixed Ireland, Netherlands

Huselid. 1995A.

Surveys of 968 large firms with more than 100 employee in 1992 and 740 in 1994

HPWS Tobin’s Q A one standard deviation increase in HPWS correlates to increased market value of $38,000 to $73,000

US

Huselid. 1995B.

Survey of 968 firms in 1992 with more than 100 employees

HPWS Profitability, Tobin’s Q

A one standard deviation increase in HPWS correlates to increased profit per employee of $3,814

US

Huselid. 1996. Surveys of 218 firms

HPWS ROA, Tobin’s Q

Positive US



Results Country

Huselid. 1997A.

Survey of 293 firms

HR policy ROA, Tobin’s Q

Positive US

Huselid. 1997B.

Survey of 702 publicly held firms with more than 100 employees and $5 million in sales

HPWS Tobin’s Q Positive US

Ichniowski. 1990.

Survey of 65 manufacturers

HR policy Tobin’s Q Positive US

Khatri. 2000. Survey of 915 large companies

HR policy Profitability Positive Singapore

Kruse. 2012. Surveys of 780 firms that ap-plied for the 100 Best Compa-nies to Work For in America list between 2005 and 2007

HPWS ROE Positive US

Lam. 1998. Survey of 235 large firms in 14 industries

HR policy ROA, market value

Positive US

Lee. 1996. Survey of 48 firms listed on the Korea Stock Exchange

HR policy ROA, ROE No correlation South Korea

Lee. 1999. Survey of 129 manufacturers in automotive parts, electron-ics, machinery and textiles

HR policy ROA Positive South Korea

Liouville. 1998.

Survey of 271 small and mid-sized industrial firms in Eastern France, 75% of which considered themselves as family firms

HR policy Profitability Positive France

Mitchell. 1989. Survey of 495 business units

HR policy ROA, ROI Mixed US

Molina. 2002. Survey of 405 firms

HR policy TSR, Tobin’s Q

Positive US, Canada

Ngo. 1998. Survey of 253 multinationals with more than 50 employees in Hong Kong

Training policy Profitability Mixed Hong Kong

Ngo. 2008. Survey of 600 firms

HR policy ROA, ROI Positive China



Results Country

Patterson. 1998.

Survey of 67 manufacturers with 60 to 1,000 employees


Patterson. 2004.

Survey of 80 manufacturing firms with 60 to 1,150 employees


Paul. 2003. Survey of 35 software firms based in Bangalore or Chennai and started before 1997

HR policy Profitability, ROI

Positive India

Richard. 2001.

Survey of 73 banks in California and Kentucky

HR policy ROE Positive US

Rodríguez. 2003.

Survey of 120 manufactur-ing firms with 100 or more employees

HR policy ROA Positive Spain

Snell. 1995. Survey of 102 single business unit firms with revenue and assets great than $10 million and at least 250 employees

HR policy ROA Positive US

Stirpe. 2009 Survey of 96 firms with 11 to 99 employees

HPWS Profitability Positive UK

Tamkin. 2008. Survey of 2,500 firms with 25 or more employees

HPWS Profitability A 10% increase in HPWS correlates to an increase in gross profits per employee of be-tween £1,083 and £1,568

UK

Thang. 2005. Survey of 137 companies in Ho Chi Minh city with at least 100 employees

HR policy Profitability Positive Vietnam

Vandenberg. 1999.

Survey of 3,570 employees at 49 life insurance firms

HPWS ROE Mixed US, Canada

Vanhala. 2006.

Surveys of 91 firms in the metal industry and retail trade between 1997 and 2000

HR policy Profitability No correlation Finland



Results Country

Watson Wyatt. 2002.

Surveys of 750 firms with at least 1,000 employees and $100 million or more in revenue or market value

HR policy TSR, Tobin’s Q

Firms with low scores on an HR policy index averaged 21% TSR over five years; medium scorers av-eraged 39%; high scorers averaged 64% years.

US, Canada, 16 European countries

Wright. 1999. Surveys of 38 refineries

HR policy Profitability Positive US

Wright. 2003. Survey of 5,635 employees at 50 largely autonomous business units of a large food service firm

HPWS Profitability Positive US, Canada

Wright. 2005. Surveys of 13,005 em-ployees at 45 business units of a large food service firm

HR policy Profitability Mixed US

Yanadori. 2014.

Survey of 17,697 non-managerial employees at 4,000 work-places

HPWS Profitability Positive Canada

Vermeeren. 2014.

A survey of 162 Dutch nursing and home care firms

HR policy Profitability Positive Netherlands

FTE=full-time employeeHPWS=high performance work systemsHR policy=human resource policyROA=return on assetsROE=return on equityROI=return on incomeROCE=return on capital employedTSR=total shareholder return*The authors did not specify the number of firms surveyed in each year


Studies of work systems proliferated in the 1980s after U.S. manufacturers realized that Japanese firms sometimes bested them on price and quality in large part due to their HR policies. As research expanded on U.S. firms, the initial focus was on discrete policies such as teamwork production methods, employee involvement in decision-making, job rotation and pay for performance. It quickly became apparent that these policies frequently worked in tandem with each other and needed to be considered as elements of a whole. Some researchers lumped together those that deal with a specific facet of HR management, such as employee compensation. One of the first studies to consider investment outcomes found positive correlations to return on assets and return on investment among 495 business units drawn from a sample of private-sector U.S. employers and a suite of alternative pay systems, including profit-sharing, gain sharing, employee stock options, employee stock ownership plans and production incentive or bonus plans.63

As the field evolved, experts pointed out the need to examine policies across all aspects of HR management. One spur was an influential 1996 study which found that five papers had used a total of 27 different variables, with a relatively low degree of overlap among them.64 An analysis published the following year warned that it could be “a recipe for disaster” if companies adopted policies that make sense in isolation but not when applied with other policies.

“Simple examples can be found in firms that invest in sophisticated performance management systems only to adopt compensation policies that provide for little meaningful economic distinction between high and low performing employees; or firms that encourage employees to work together in teams, but then provide raises based on individual contributions.”65

HR Policy BundlesSince then many researchers have attempted to identify the most effective sets of policies, which often are referred to as “bundles.” As a 2005 paper said: “it is not practices per se that make the difference but the degree to which they align with each other to create meaningful ‘bundles’ of practice. Various studies have found that adoption of single practices does not deliver the same improvement of results.”66

The 2003 British Task Force on Human Capital Management mentioned above commissioned surveys to identify the use of bundles both in the UK and in 13 other countries:


“We concluded that there is no single set of HCM practices widely accepted as ‘best practice’ applicable to all organisations, nor agreement on a set of universally relevant indicators. However, there is a reasonable consensus on the range of practices that might be relevant dependent on the particular circumstances of the organisation and the business strategy it is following.”67

There has been less agreement on what to call such bundles and what each might encompass. Some experts have used generic terms such as HR management while others attempt to identify what have been termed “high-performance work systems.” A 2013 paper summarized the latter school of thought:

“A number of different labels have been used to describe research on the relationship between work and employment practices and performance, including high performance work systems, high commitment work systems, high involvement work systems and high performance human resource management. The common finding emerging from these studies is that achieving and sustaining high levels of performance requires a combination of workplace innovations to leverage employees’ knowledge and ability to create value and coordinate their efforts to work together. That, in turn, produces and sustains a positive workplace culture and practices. While the specific practices need to be tailored to fit different industries and occupations, they generally include selection, training, mentoring, incentives, knowledge-sharing, engaging front-line workers in operational decisions, and partnership based labor-management relations and other shared decision making mechanisms to address broader issues. These practices were found to be most effective when implemented together and in concert with new capital or technological investments.”68

Our survey of the field encompasses all the approaches. For the sake of simplicity we have chosen to use the term human resource (HR) to characterize the literature. A sample of the dozens of HR policies assessed by researchers can be seen in Table 4.


Table 4: Types of HR Policies

Compensation and Benefits

Pay for Performance

Formal Appraisal for Pay

External Pay Equity/Competitiveness

Incentive Compensation

Comprehensive Benefits

Profit or Gain Sharing

Group-Based Pay

Pay for Skills/Knowledge

Employee Stock Ownership

Bonuses or Cash for Performance

Equitable Pay Processes

Public Recognition/Nonfinancial Rewards

Job and Work Design

Decentralized Participative Decisions

Project or Other Temporary Work Teams

Job Analysis

Job Rotation/Cross Functional Utilization

Self-Managed Work Teams (Quality Circles)

Greater Discretion and Autonomy

Job Enlargement and Enrichment

Broad Task Responsibilities

Flexible Work Schedule

Training and Development

Training Extensiveness

Use of Training to Improve Performance

Training for Job or Firm Specific Skill

Training for Career Development

Evaluation of Training

Cross-Functional or Multiskill Training

New Employee Training and Orientation

Recruiting and Selection

Hiring Selectivity or Low Selection Ratio

Specific and Explicit Hiring Criteria

Multiple Tools Used to Screen Applicants

Employment Tests or Structured Interviews


Planning Selection Processes and Staffing

Matching Candidates to Firm Strategy

Innovative Recruiting Practices

Employee Relations

Job Security/Emphasis on Permanent Jobs

Low Status Differentials

Complaint or Grievance Procedure

Measurement of Employee Relations Outcomes

Employee Opinion and Attitude Surveys

Labor Union Collaboration

Social and Family Events and Policies

Diversity and Equal Employment Opportunity

Communication

Formal Information Sharing Program

Employees Receive Market, Firm Performance, or Strategic Information

Employee Input and Suggestion Processes

Frequent/Regular Meetings with Employees

Performance Management and Appraisal

Appraisals Based on Objective Results/Behaviors

Appraisals for Development/Potential

Frequent Performance Appraisal Meetings

Employees Involved in Setting Appraisal Objectives

Written Performance Plan With Defined Objectives

Multisource Feedback and Peer Appraisal

Appraisal Based on Strategic or Team Goals

Promotions

Promotions From Within

Promotions Objectively Based on Merit

Career Planning

Promotion Opportunities (e.g., frequency)

Career Paths and Job Ladders

Succession Planning

Turnover, Retention, and Exit Management

Taken from: Posthuma. 2013. The authors found that these 61 policies appeared a total of 2,042 times in 181 peer-reviewed academic and practitioner studies of high performance work systems published between 1992 and 2011. These included studies that examined cor-relations to investment outcomes as well as those that looked at other variables.


Some of the early efforts to correlate HR policy bundles to investment outcomes were undertaken in a series of papers in the 1990s by two US academics, Mark Huselid, then at Rutgers University, and Brian Becker of the State University of New York at Buffalo. They largely followed the template Huselid drew up in an initial 1995 study of 13 HR factors using a survey of 968 publicly traded U.S. firms with more than 100 employees. He found positive correlations between an index of these factors and both Tobin’s Q and gross returns on capital. While much of the research in the field stops with such regression analysis, Huselid went on to estimate that a one-standard deviation increase in the index was associated with an $18,641 gain in market value per employee and a $3,814 increase in profits per employee.69 These were intended as an approximation rather than a hard-dollar estimate given the underlying assumptions. For example, Huselid held other variables at their means and “arbitrarily” assumed that the returns accrued over a five-year period at an eight percent discount rate. This kind of analysis also provides only implicit estimates of the cost companies incur in adopting such HR policies. Still, his point was to offer some estimate of the practical effects of increasing use of these practices. He and Becker came to similar conclusions in five more studies using various surveys of U.S. companies over the subsequent several years.70 Many others have created similar indices of HR factors as a way to measure links between bundles of policies and performance. One of the most sophisticated efforts came in a four-year British project called People and the Bottom Line. Researchers created a model with 40 human capital measures that were the basis for a survey of 2,905 British employers with 25 or more employees, including 2,500 private-sector ones. Responses were organized into four indices including training; employee involvement in decisions; HR strategy; and recruitment. Regression analysis was used to look for correlations to gross profits per employee, operating profit per employee and profit margins per employee. The authors found only weak correlations to each separate policy bundle, but strong ones for the four indices taken as a whole.71 They then used the private-sector firms’ reported financial data to estimate the gains to companies with higher overall index scores. The authors described their findings as follows:

“The size of the effects are also of note and provide, in tangible terms, a sense of the relationship between the index and the organisation’s performance. The results imply that if a business increases its investment by the equivalent of increasing its combined index score by one (around 10 per cent), this would equate to:

• an increase in gross profits per employee of between £1,083 and £1,568.• an increase in operating profit per employee of between £1,139 and £1,284.• an increase in profit margins per employee of between 1.19 per cent and 3.66 per cent (i.e. the ratio of profit over sales).”72


Still, as with Huselid and others who translate correlations into actual profit numbers, the assumptions involved in such estimates make them less robust than the specificity of the figures might suggest.

Indices of HR policy have been used extensively in practitioner literature, which often was designed to support company decision making and therefore did not always present underlying details on data and methodology as is common among academic studies. For example, a 2002 study by Watson Wyatt (now part of Towers Watson) started with 1999 surveys of HR policies at 400 publicly traded U.S. and Canadian companies. It then conducted regression analyses to search for correlations to market value, three- and five-year total shareholder returns and Tobin’s Q. It concluded that a bundle of 30 practices correlated to an average 30 percent increase in market value. The firm conducted a similar survey the following year of 250 firms in 16 European countries and came to a comparable conclusion, identifying 19 HR policies associated with a 26 percent increase in market value. Finally in 2001 it repeated the North American survey with 500 companies, 51 of which had participated in the 1999 survey. The two data sets were merged into one with more than 750 companies in the United States, Canada and Europe with at least three years of shareholder returns, 1,000 or more employees and a minimum of $100 million in revenues or market value. The study then created what it called a Human Capital Index (HCI) based on the HR factors identified. It stated that “[t]he higher a company’s HCI score, the higher its shareholder value. In other words, the better an organization is doing in managing its human capital, the better its returns for shareholders. We broke the companies into three groups based on their summary HCI scores. Those in the low group averaged a 21 percent five-year return. The medium group averaged 39 percent. Those with high HCI scores returned 64 percent over five years.”73

A 2009 paper used an index to evaluate the relationship of HR to credit risk. What the authors termed as an Employee Relations Index drew on HR factors identified by Huselid and others. They employed it to analyze 2,265 bonds issued by 568 U.S. firms between 1995 and 2006 and found that companies with higher scores had lower cost of debt, lower credit risk, higher credit ratings and lower stock volatility. For example, a one-point increase of the index score was associated with a decrease in the annual yield spread of two to four basis points.74

Over the years, a handful of the 56 studies have found mixed correlations between HR policies and financial performance, and two found none at all. One of the latter, published in 1996, surveyed 48 manufacturing firms listed on the Korea Stock Exchange, asking senior managers or executives about ten HR policies relating to employee participation


as well as questions about the company’s HR strategy. The paper used statistical analyses to assign firms to four groups according to the way firms employed various HR policies. Regressions turned up no associations to prior three-year return on assets or three-year return on equity.75

Nine other papers from a variety of countries uncovered positive correlations for some outcomes and none for others. The largest was based on two years of survey data submitted by 633 firms that were among those included in the Dow Jones Sustainability Index.76 The author drew up four different indices drawing on HR bundles studied in other papers. Only one index, comprised of policies on training and skills gap measurements, showed positive associations to two financial outcomes, return on assets and Tobin’s Q. However, it did not correlate to prior three-year total shareholder returns. The other three indices did not correlate to any of the financial outcomes.77 The author’s conclusion: “Taken as a whole, the important message that emerges from the analyses so far is that some, but not all, elements of human capital management display a relation with firm valuation.78

The “Black Box” QuestionAlthough these and the other studies in Table 3 establish links between HR policies and financial performance, researchers are still working to document and measure the complex intervening organizational processes between the two. This is frequently referred to

the materiality of human capital to corporate financial ......human capital research has been...

Documents